There is considerable interest in identifying biomarkers to aid in the early identification of individuals at risk for developing mild cognitive impairment (MCI) and Alzheimer's disease (AD). The prevalence of AD is expected to double every 20 years, making its eventual burden on the healthcare industry of serious concern if this course is not altered. Changes in white matter integrity and consequent disruption of the connectivity between regions critical for cognitive functioning is recognized as an important underlying pathophysiology in MCI and AD. Diffusion Tensor Imaging (DTI) is an imaging technique that provides a microstructural analysis of white matter integrity and is sensitive to white matter changes often undetected by other imaging modalities. Currently, only a few DTI studies have been conducted with healthy at-risk individuals (defined by a family history of AD plus one or both Apolipoprotein E 54 alleles). These studies have been cross-sectional and have most often used only one index of diffusion, fractional anisotropy (FA). A recent cross-sectional investigation showed increased axial diffusivity (DA) and radial diffusivity (DR) in an at-risk, cognitively intact older group compared to a demographically matched group without risk factors. These diffusion changes were observed over widespread regions of the cerebral white matter and were present in the absence of changes in FA. At present, no study has examined longitudinal DTI measurements in persons at risk for AD. As part of an NIH- funded longitudinal project (NIA R01-AG022304; S. Rao, Principal Investigator), persons ranging in age from 65-89 and at varying risk for developing AD have undergone task-activated and resting-state functional magnetic resonance imaging (fMRI) and DTI examinations at baseline, 18 months, and 58 months. With the requested support, longitudinal DTI white matter changes over the 18- and 58-month follow-up intervals as a function of AD risk will be investigated using multiple DTI parameters (FA, DA, and DR). It is hypothesized that individuals at high risk for AD will show accelerating increases in DA and DR over the 18- and 58-month follow- up assessments. In addition, it is predicted that baseline DTI indices will successfully predict individuals who show subsequent memory decline over the 18- and 58-month intervals. Finally, the DTI acquisition acquired at the 58 month follow-up will be examined using a probabilistic fiber tracking approach to investigate white matter tracks in a semantic memory activation network that overlaps the default mode network (DMN), a neural system compromised in both AD and MCI. It is hypothesized that individuals at-risk for AD will also show disruption of the white matter connections serving this network. The proposed investigation is consistent with NIA's mission to extend the healthy, active years of life by determining if DTI is a useful biomarker in predicting memory decline and conversion to MCI and AD in pre-symptomatic individuals.